What is the genomic architecture of a quantitative trait? Despite substantial effort to answer this question in humans and model organisms, we remain far from understanding how many genes, gene-gene interactions, and gene-environment interactions underlie most polygenic traits, such as human disease. Global gene expression studies, which treat each transcript in the genome as a quantitative trait, have provided crucial insights into the genetic basis of trait differences between individuals. In particular, a cross of the BY lab strain and the RM wine strain of the budding yeast Saccharomyces cerevisiae that was done by the lab of Leonid Kruglyak, where I am presently a postdoctoral fellow, has illuminated the genetic complexity underlying expression differences between two individuals. However, even within this cross, the genetic variance for the majority of the transcripts in the genome remains incompletely mapped, with the summed effects of detected linkages often explaining only a small fraction of a transcript's expression variance. I am developing a new method that, for many polygenic architectures, will facilitate the mapping of all linkages in the genome that underlie a transcript difference between two yeast strains in a single environment, potentially with a gene-level mapping resolution. This approach exploits aspects of the recently developed Synthetic Genetic Array (SGA) technology to create extremely large pools (~10'^5 to 10'^7) of recombinant MATa haploids from a single cross. Bulk segregant analysis (BSA) on these large populations, which can be done by using parents that harbor translational fusion fluorescent reporters and cell sorting/recapture on the segregant pool, will facilitate the mapping of the genomic architecture of target transcripts. Once working, this approach can be extended to multiple environments, other selectable traits (e.g. drug resistance), and new backgrounds. AIM 1: To develop a robust metholodogy for mapping the genomic architecture of expression quantitative traits in large pools of segregants. AIM2: To apply this method to 25 transcripts that have previously been shown to exhibit heritable variation across segregants of a BY X RM cross in a glucose-limited environment. AIM3: To validate the genomic architecture of one transcript by doing all necessary allele replacements in both the BY and RM backgrounds. Public Health Relevance: Many diseases are influenced by multiple genes, with the number of carried risk alleles varying from person- to-person. Understanding how many genes contribute to risk for a particular disease remains a major challenge for medical genetics. The experiments I propose on yeast gene expression can provide critical information about how many genes underlie trait variation, such as disease risk.